Measure AI traffic in Matomo – Guide

ChatGPT, Claude, Perplexity – they send visitors to websites. In Matomo, this traffic ends up in the “Referral” channel, unsorted and invisible. This guide shows three ways to cleanly capture, analyze and continuously monitor AI traffic in Matomo – without external tools, in a privacy-compliant way and with historical data.

1. Matomo vs. GA4: what is different?

Anyone familiar with the GA4 guide will notice one key difference: Matomo does not have custom channel groups. GA4 solves the problem through a rule set that classifies future traffic. Matomo takes a different approach, and in some respects a more elegant one.

2. Why measure AI traffic separately?

AI systems are the new discovery layer: they read content, evaluate sources and provide concrete recommendations. If your site appears as a reliable source in ChatGPT or Perplexity, you get traffic. If it is not mentioned, you do not exist for a growing share of users.

In Matomo, this traffic – just like in GA4 – ends up in the “Referral” channel, mixed together with newsletter clicks, partner links and spam. The result: you cannot see whether AI systems are sending relevant visitors, which pages benefit from it and how well these users convert.

3. Method 1: Create a segment (recommended)

Segments are the fastest and most flexible solution. They apply retroactively, are available in every report and can be adjusted at any time – without data loss.

1. Click the segment selector at the top (dropdown next to the calendar).
2. Select + Add segment → the segment editor opens.
3. Name the segment: AI Traffic.
4. Under “Conditions”, click + Add condition.
5. Select the dimension: Referrer URL (under “Visits”).
6. Operator: matches regular expression.
7. Value: paste in the regex from section 4.
8. Click Save. Matomo calculates the segment retroactively – depending on the amount of data, this may take a few minutes.

4. The AI regex: complete list of all sources

This regex covers all relevant AI platforms: ChatGPT, Claude, Gemini, Perplexity, DeepSeek, Microsoft Copilot, Grok, Mistral, Hugging Face and more than 80 other sources. It is identical to the GA4 version and can be used directly in Matomo.

5. Method 2: Configure channel attribution (from Matomo 4.x)

Channel attribution complements the segment: while segments are used for analysis, channel attribution permanently classifies traffic as a separate channel – visible in the report Referrers → Channels.

1. In the Matomo backend, open the desired website under Administration → Websites.
2. Select the Channel attribution tab.
3. Click + Add new rule.
4. Enter the channel name: Artificial Intelligence.
5. Condition: Referrer URL → matches regular expression → paste in the regex.
6. Place the rule above “Referral” in the order (drag & drop).
7. Save. From now on, AI traffic will be classified as a separate channel.

6. Method 3: Custom Dimension via Tag Manager

For granular analyses – for example, to see which specific AI platform sent a visitor – a Custom Dimension is recommended. This is set on every page view via the Matomo Tag Manager.

Requirement

Matomo Tag Manager must be enabled and the container must be embedded on the website. Check this in the Matomo backend under Tag Manager → Container.

Create JavaScript tag

1. Create a new tag of the type “Custom HTML”.
2. Trigger: “Page View” (All Page Views).
3. Insert the following JavaScript code:

(function() {
  var ref = document.referrer || '';
  var aiRegex = /chatgpt\.com|chat\.openai\.com|claude\.ai|perplexity(?:\.ai)?|gemini\.google\.com|deepseek\.com|copilot\.microsoft\.com|grok\.com|(?:\w+\.)?mistral\.ai|huggingface\.co|you\.com|phind\.com|duck\.ai|felo\.ai|coze\.com/;

  if (aiRegex.test(ref)) {
    // Extract the specific platform
    var platform = 'Other AI';
    if (/chatgpt\.com|chat\.openai\.com/.test(ref)) platform = 'ChatGPT';
    else if (/claude\.ai/.test(ref)) platform = 'Claude';
    else if (/perplexity/.test(ref)) platform = 'Perplexity';
    else if (/gemini\.google\.com/.test(ref)) platform = 'Gemini';
    else if (/deepseek\.com/.test(ref)) platform = 'DeepSeek';
    else if (/copilot\.microsoft\.com/.test(ref)) platform = 'Copilot';
    else if (/grok\.com/.test(ref)) platform = 'Grok';
    else if (/mistral/.test(ref)) platform = 'Mistral';
    else if (/huggingface\.co/.test(ref)) platform = 'HuggingFace';

    // Set Custom Dimension 1 (adjust dimension ID if needed)
    _paq.push(['setCustomDimension', 1, platform]);
  }
})();

1. Under Administration → Custom Dimensions → Visits, create a new dimension: Name AI Platform, Scope Visit. Adjust the assigned ID (e.g. 1) accordingly in the code above.
2. Save the tag and publish the container.
3. From now on, the dimension is available in reports under Visitors → Custom Dimensions.

7. Analyze AI traffic in reports

As soon as the segment is active, it is available in every Matomo report as a filter. The most important reports for AI traffic analysis:

Activate segment

The segment filters the entire report immediately. All metrics – visits, time on site, bounce rate, goals – then refer only to sessions that came via AI platforms.

8. AI vs. Paid: set up channel comparison

Starting with version 4.x, Matomo offers the segment comparison feature: two segments are displayed side by side in one report. This makes it possible to compare AI traffic directly with Paid Search.

Create second segment: Paid Search

1. Open the segment editor → create a new segment: Name Paid Search.
2. Condition: Channel → is equal to → Paid Search.
   Alternatively: Referrer type → is equal to → Search engines (paid).
3. Save.

Activate comparison

1. In the segment dropdown, select the first segment AI Traffic.
2. Click “+ Add segment” and add Paid Search.
3. Matomo displays both segments side by side – with all metrics from the active report.
4. Optional: add a third segment Organic Search for the full channel comparison.

9. Export data & automate

Manual export

Matomo exports the current report including the active segment in various formats. CSV is suitable for Excel analysis, PDF for client reports.

Automation via Reporting API

The Matomo Reporting API allows fully automated data queries with segment filtering. Example URL for the page report with AI segment:

https://[YOUR-MATOMO-URL]/index.php
  ?module=API
  &method=Actions.getPageUrls
  &idSite=1
  &period=month
  &date=last3
  &format=CSV
  &token_auth=[YOUR-TOKEN]
  &segment=referrerUrl=~chatgpt.com,
           referrerUrl=~claude.ai,
           referrerUrl=~perplexity.ai

Scheduled reports by email

Matomo can send reports automatically by email – daily, weekly or monthly, as PDF or CSV. The segment can be stored as a filter in the report.

10. Multi-Site: one setup for all websites

Matomo offers a decisive advantage here over GA4: in one Matomo instance, segments can be created globally and used simultaneously across multiple websites.

Create global segment

1. Open the segment editor → create a new segment.
2. When saving, select the option “All websites” (instead of a specific website).
3. The segment is now available in every website within the Matomo instance.

Channel attribution: set up per website

Channel attribution must be configured separately for each website – it is website-specific. With the prepared regex, however, this takes less than 3 minutes per website.

API-based rollout (for many websites)

Using the Matomo Admin API, segments and channel rules can be created programmatically and rolled out across any number of websites. This is useful for agencies with 10+ websites in one Matomo instance.

11. Frequently asked questions